The update explicitly embraces intellectual freedom within defined safety boundaries, allowing discussion of controversial topics while maintaining restrictions against concrete harm.
OpenAI Updates Model Spec to Better Balance User Freedom with Safety Guardrails
This was originally published on post